Thermal imaging of rice grains and flours to design convolutional systems to ensure quality and safety. (March 2021)

• Record Type: Journal Article
• Authors: Estrada-Pérez, Leidy V.; Pradana-López, Sandra; Pérez-Calabuig, Ana M.; Mena, María Luz; Cancilla, John C.; Torrecilla, José S.
• Is Part Of: Food control. Volume 121(2021)
• Publication Date: 2021-03
• Subjects: Thermography -- Rice grains characterization -- Rice flours characterization -- Convolutional neural network -- Food quality control -- Adulteration; Food -- Quality -- Periodicals; Food -- Analysis -- Periodicals; Food handling -- Periodicals; Food industry and trade -- Quality control -- Periodicals; Aliments -- Industrie et commerce -- Qualité -- Contrôle -- Périodiques; Aliments -- Qualité -- Périodiques; Aliments -- Analyse -- Périodiques; Hygiène alimentaire -- Périodiques; Food -- Analysis; Food handling; Food -- Quality; Periodicals; Electronic journals; 664.07
• DOI: 10.1016/j.foodcont.2020.107572
• Languages: English
• ISSNs: 0956-7135

Abstract

Abstract: In this work, a thermographic camera and intelligent algorithms have been used to classify five different types of rice ( Oryza sativa L.) in grain or flour format and to detect mixtures of different rice types which act as adulterated samples. For this purpose, more than 63, 000 thermographic images of pure rice (Japonica and Indica) and their mixtures were used. These images have been used to train and validate a system based on deep learning (convolutional neural networks) to carry out their classification and detect adulterations. The combination of cognitive modelling and thermographic analysis can classify the different types of rice and detect potential adulterations with accuracies above 98%. These promising results make this tool viable for fraud detection, food quality control, and public safety. Highlights: Thermal evolution of rice captured via thermographic camera. Thermography and convolutional neural networks combined to fight adulteration. Five types of rice classified virtually perfectly through deep learning models. Fraudulent mixtures of rice detected approximately 99% of cases. Quality control of rice enabled for real-time analysis and evaluation.